Preamble DA1855

From TekWiki
Jump to navigation Jump to search
Preamble Instruments DA1855
100 MHz differential amplifier
Preamble DA1855

Produced from (?) to (?)

Manuals
(All manuals in PDF format unless noted otherwise)
Manuals – Specifications – Links – Pictures

The Preamble (LeCroy) DA1855 is a 100 MHz differential amplifier, an updated, stand-alone version of the Tektronix 7A13.

The principal designer was John Addis, who contributed the following background:

Overload is an allowable input without destruction. Overdrive, on the other hand, occurs when input signal causes the amplifier to become non-linear. This is apparent in how the amplifier recovers to its linear operation. By definition, all amplifiers suffer from inaccuracies after overdrive, let alone overload. Overdrive occurs with smaller inputs than overload and occurs when the amplifier behavior becomes obviously distorted in recovery. Differential comparators are designed to minimize that recovery time. Usually, thermal effects are most relevant to accurate measurements.

Differential comparators are, in historical order, the Tektronix Z plugin, W, 10A1, 7A13, HP 1803A, the Preamble 1850, 1855, 1850A, 1855A, and Lecroy DA1855A. These amplifiers were designed to accommodate large overdrives without bad thermals. They were successful in varying degrees, and the technology used changed over time.

The Z plugin (introduced in 1960) was designed by Charlie Rhodes of TV fame. The Z used vacuum tube technology almost exclusively. There were a few transistors, but they were not in the signal path. Both input devices were separately bootstrapped. Overload and common mode dynamic range were a very substantial ±100V! Bandwidth was an impressive 13 MHz.

The W plugin (introduced in 1965) used 8056 Nuvistors in the input and three 8416 dual triodes along with transistor amplifiers. Overdrive was handled with a single disconnect diode per side. One stage has feedback. Overload and common mode dynamic range is about ±12 V. The W was designed by John Horn, who was a large man. The very large knob for coarse comparison voltage was therefore sometimes called the John Horn knob.

The 10A1 (introduced in 1965) was a much more clever design using the collector-base and emitter-base junctions of the same transistor to disconnect large signals with minimal thermals. The circuit is quite sophisticated. The input used individually bootstrapped Nuvistor buffers. Common mode dynamic range was about 7 V. Bandwidth was 45 MHz at ≥5 mV/div, 35 MHz at 1 mV/div in a 647 mainframe. The 10A1 was designed by Les Larson.

The Tektronix 7A13 (introduced in 1969) was the first differential comparator to make use of JFET inputs. The 7A13 used more conventional brute force overdrive circuitry, but the JFET inputs, large overload and common mode dynamic range, and the compatibility with 7000 series made it far more popular than the 10A1. Overload and common mode dynamic range is about 11 V, bandwidth is 100 MHz with a 2-pole bandwidth limit filter at 5 MHz to get rid of noise. The 7A13 was designed by Bill DeVey.

The HP 1803A (introduced in 1972) was largely a copy of the Tektronix 10A1 with essentially the same input circuit and specifications except that the 1803A used JFET inputs with consequent better stability. It was designed by HP’s John Cardon.

The 11A33 (introduced in 1989) used overdrive circuitry of the M377 IC developed for the 11A32, 11A34, 11A52 plugins. This is a feedback amplifier which changes its configuration in overdrive but maintains a closed feedback loop. Its monolithic construction made it the preferred method of fast overdrive recovery.

However, the common mode dynamic range of the M377 is only about ±3.5 V. So the 11A33 uses a separate IC in front of its M377 with totem-poled input devices to obtain about 8.5 V of overdrive and common mode dynamic range using the same 15 V IC process. Bandwidth is 150 MHz with four pole filters at 100 MHz and 20 MHz. The 11A33 was the first differential amplifier with calibrated true differential offset. Pat Quinn designed the input IC. The 11A33 had a redesign shortly after introduction. I think the redesign used the later SHPi IC process which had p-channel JFETs, but the same 15 V breakdown. The redesign was done by Rich Huard. Within the limitation of 8.5 V overload and common mode dynamic range, the 11A33 had, I think, the best overdrive recovery ever, but with limited common mode dynamic range.

Preamble Instruments was a small start-up eventually employing 28 people. Products included the 1820, 1822, 1850, 1855, 1820A, 1822A, 1850A and 1855A differential amplifiers and several active and passive differential probes. (The 1800 series A versions had μP-controlled operation designed by Stan Sasaki and oven heater that only ran with power on). The 1850, 1850A, 1855, 1855A (introduced in 1996) used input circuitry based on the M377. The design used all discrete circuitry and therefore could not quite match the 11A33 for thermals, but its 16 V overload and common mode dynamic range, true differential calibrated offset, along with 10 MHz, 1 MHz, and 100 kHz BWL filters made it very popular. LeCroy Corporation bought Preamble Instruments in October 1997. It changed the model name to DA1855A. LeCroy is now known as TELEDYNE LECROY. The DA1855A is still in their catalog as of this writing in November 2018. Twenty-two years is a good lifetime. I designed the M377 and the 1800 series differential amplifiers.

In general, any instrument that has a slideback (aka comparison) voltage will behave in overdrive better than other amplifiers. The Tektronix 1A7A and 7A22 were not very fast (1 MHz), but used feedback circuitry that recovered very well within their bandwidth. Their available offset voltage was differential, allowing true differential performance over a limited range (1 V at the most sensitive range), but the offset was not calibrated. The 1A7A (designed by Thor Hallen) and the 7A22 (designed by Val Garuts) used essentially the same circuit. Exactly who came up with the brilliant input circuit is lost to history. The 1A7A came out first, but Thor had worked as evaluation engineer for Val Garuts. Val does not remember who created the circuit, and Thor died in 2002.

I designed the 1A7 (using Nuvistors) and will take credit for the combination of features of the 1A7A and 7A22, but the circuitry of the 1A7A and 7A22 is far superior and used the much more stable JFETs input devices.

John Addis November 16, 2018

Links